Framework for Large- Scale Communica5on and Power Co- Simula5on

Similar documents
Energy Storage Systems and Role for the US Grid

Using Demand Response Programs to Benefit the. PtikJ Patrick J. Oshie, Ohi Commissioner Washington Utilities & Transportation Commission

PJM LMP Market Overview

PJM Overview and Wholesale Power Markets. John Gdowik PJM Member Relations

Dr. Raheem Beyah Georgia Tech. Cyber-Physical Modeling & Simulation for Situational Awareness (CYMSA)

High Performance Computing for Wireless Telecom Research

The State of the Electrical Grid in Washington State. Michael Pesin, PMP, P.E. Seattle City Light

SCADA System Overview

Enabling 24/7 Automated Demand Response and the Smart Grid using Dynamic Forward Price Offers

Distributed Generation and the Impact on a Distribution System. Brian Lassiter Technical Sales Engineer

Panel Session: Lessons Learned in Smart Grid Cybersecurity

Demand Response in Capacity and Electricity Markets: What Role Can and Should It Play?

SCE s Transactive Energy Demonstration Project. GWAC Workshop Bob Yinger December 10-11, 2013

Presentation for The National Commission for Energy State Regulation of Ukraine

COMMENTS OF THE SOLAR ALLIANCE NEW JERSEY INTERCONNECTION RULES APRIL 29 TH, 2011

PHASOR MEASUREMENT UNIT (PMU) AKANKSHA PACHPINDE

Enabling the SmartGrid through Cloud Computing

Advanced Distribution Grid Management for Smart Cities

Smart Storage and Control Electric Water Heaters. Greg Flege, Supervisor, Load Management Midwest Rural Energy Conference March 3, 2011

Measurement of V2oIP over Wide Area Network between Countries Using Soft Phone and USB Phone

Hybrid Simulation von Kommunikationsnetzen für das Smart Grid

CERTS Microgrid Research and Lessons Learned

Modeling of GE Appliances: Cost Benefit Study of Smart Appliances in Wholesale Energy, Frequency Regulation, and Spinning Reserve Markets

IEEE-Northwest Energy Systems Symposium (NWESS)

Two-Settlement Electric Power Markets with Dynamic-Price Contracts

Synchronized real time data: a new foundation for the Electric Power Grid.

Modelling electric vehicle demand in London using the DCE platform

Public Service Co. of New Mexico (PNM) - PV Plus Storage for Simultaneous Voltage Smoothing and Peak Shifting

The Economics of Cisco s nlight Multilayer Control Plane Architecture

Distribution System Analysis Tools for Studying High Penetration of PV with Grid Support Features

Cost-Benefit Analysis of Demand Response Programs Incorporated in Open Modeling Framework

Testing Intelligent Device Communications in a Distributed System

Next Generation Distribution Management Systems (DMS) and Distributed Energy Resource Management Systems (DERMS)

Introduction to Ontario's Physical Markets

Operating the power system while enabling more renewables : which solutions?

Georgia Tech ARPA-E: Energy Internet

SONET and DWDM: Competing Yet Complementary Technologies for The Metro Network

How to deal with a thousand nodes: M2M communication over cellular networks. A. Maeder NEC Laboratories Europe andreas.maeder@neclab.

Introduction to MATLAB Gergely Somlay Application Engineer

Energy Smart Data Center Phase-III R&D. Andrés Márquez. Pacific Northwest National Laboratory

Asset Management Challenges and Options, Including the Implications and Importance of Aging Infrastructure

A Scalable Network Monitoring and Bandwidth Throttling System for Cloud Computing

Cyber Infrastructure for the Smart Grid

Cybersecurity Risk Assessment in Smart Grids

Schneider Electric DMS NS. Company Profile. Commercial Documentation

Convergence of Wholesale And Retail Markets: The Texas Experience

Why Smart Water Networks Boost Efficiency

Smart Grids. MIECF Conference April 2011

Advanced Data Management and Analytics for Automated Demand Response (ADR) based on NoSQL

ERCOT Monthly Operational Overview (March 2014) ERCOT Public April 15, 2014

Internet Protocol (IP)/Intelligent Network (IN) Integration

PJM Overview of Markets. Georgian Delegation PUCO Office April 11, 2013

Realization of control center HMIs by using IEC and CIM data bases for communication and data handling

Office of Electricity Delivery & Energy Reliability ANALYSIS AND REPORTING OF METRICS AND BENEFITS FOR ARRA SMART GRID PROJECTS

Overview and Comparison of Demand Response Programs in North American Electricity Markets

Mr. Apichon Witayangkurn Department of Civil Engineering The University of Tokyo

technology standards and protocol for ip telephony solutions

Next Generation Grid Data Architecture & Analytics Required for the Future Grid

Demand Response Management System Smart systems for Consumer engagement By Vikram Gandotra Siemens Smart Grid

Data Analysis with MATLAB The MathWorks, Inc. 1

White Paper. Convergence of Information and Operation Technologies (IT & OT) to Build a Successful Smart Grid

DESIGN FOR MONITORING AND OPTIMIZATION OF POWER DEMAND FOR WIRELESSLY COMMUNICATING ELECTRIC LOADS

StreamStorage: High-throughput and Scalable Storage Technology for Streaming Data

Steve Lusk Alex Amirnovin Tim Collins

Smart grid security analysis

Advanced Electricity Storage Technologies Program. Smart Energy Storage (Trading as Ecoult) Final Public Report

Strategic Microgrid Development for Maximum Value. Allen Freifeld SVP, Law & Public Policy Viridity Energy

Integration Capacity Analysis Workshop 11/10/15 California IOU s Approach

Evolution of Control for the Smart Transmission Grid

Con Edison Demonstration Project DE-OE Customer Sited Resource Management Tools

PARALLEL & CLUSTER COMPUTING CS 6260 PROFESSOR: ELISE DE DONCKER BY: LINA HUSSEIN

Il Progetto INTEGRIS. Risultati e nuove prospettive per lo sviluppo delle Smart Grid October 2nd 2013 Brescia

Robust Adaptive Topology Control

Isolines: Energy-efficient Mapping in Sensor Networks

Gaming as a Service. Prof. Victor C.M. Leung. The University of British Columbia, Canada

DMS - Breakthrough Technology for the Smart Grid

ALSTOM Energy Management Business. Challenges related to Smart Energy Eco Systems

C-DAX: A Cyber-Secure Data and Control Cloud for Power Grids C-DAX Consortium

A Systems of Systems. The Internet of Things. perspective on. Johan Lukkien. Eindhoven University

ARRA Grant Case Studies SMUD s Smart Grid Program

Developing Scalable Smart Grid Infrastructure to Enable Secure Transmission System Control

1. Summary. electric grid, strengthen saving programs sponsored by utilities. The project

Dynamic Network Resources Allocation in Grids through a Grid Network Resource Broker

Modeling of PV Based Distributed Generator Systems with Diverse Load Patterns

What Applications Can be Deployed with Software Defined Elastic Optical Networks?

Utility Analytics, Challenges & Solutions. Session Three September 24, 2014

Venteea A Smart Grid demonstrator for the DRES integration on the MV distribution networks.

Spring Final Project Report

Transcription:

1 Framework for Large- Scale Communica5on and Power Co- Simula5on Jason C. Fuller Pacific Northwest Na6onal Laboratory ISGT 2014-02/21/2014 jason.fuller@pnnl.gov PNNL- SA- 101044

Why do we need to co-simulate? Smart grid brings information and communication technologies together with power systems Sensors and equipment gather information Information is processed locally or centrally Decisions are made based on this information But before deploying new technologies, it is important to understand: What is the performance of a given technology? How will new technologies interact with existing technologies? Will communication latency affect the controllability or observability? What are my communication system requirements to support an application? Can applications share network bandwidth?

Building an integration framework Traditionally, power grid and communication network domains have not resided within a single simulator with relatively equal consideration to the complexity of each. A number of very powerful, domain-specific tools exist: Transmission (PSLF, Powerworld, DSATools, PST, etc.) Distribution (WindMil, SynerGEE, CYMDIST, OpenDSS, GridLAB-D, etc.) Telecommunications (OPNET, NetSim, ns-2, ns-3, OMNet++, etc.) We do not need to recreate these tools Re-use existing simulators Libraries of models already exist Most are well validated Integrate and enjoy!!

Scalability and Co-Simulation Co-simulation allows for expansion of capabilities with minimal investment Allows for re-use of existing software AND models Enables multi-scale modeling and simulation required for understanding TC2 FNCS is a framework for integrating simulators across multiple domains Framework for Network Co-Simulation (FNCS pronounced Fee-nix ) Developed for HPC applications across multiple platforms PowerWorld Transmission (GridPACK) Wholesale Markets (Matpower) GridLAB-D GridLAB-D Communications (ns-3) GridLAB-D Connected In Development FNCS EnergyPlus EnergyPlus EnergyPlus Distribution (GridLAB-D) Buildings (EnergyPlus) Retail Markets (GridLAB-D)

Intended uses? Distribution and Communications Sensor data and control (VVO, inverters, reconfiguration, etc.) Demand response and retail markets Transmission and Communications Wide Area Control (and Protection) Phasor Measurement Unit data collection and control Communication pathways and redundancy Transmission, Distribution and Communications Trade-offs of distributed versus centralized controls Hierarchical controls / reconfiguration during communication loss Transmission, Distribution, Markets and Communications Transactive energy/ancillary markets (with distributed resources) Integration of wholesale and retail markets Visualization Generate simulated data sets for experimentation

FNCS Framework Architecture FNCS API consists of three (minimal) components Time management / synchronization Communication interface between broker and simulator(s) Message delivery and serialization Programmers need to use the components All other components are hidden to ease the programming requirements FNCS is programmed in C++, and interfaces for C, Java, Fortran are provided with FNCS distribution

Demand Response/Real-Time Pricing Example RTP, double-auction, retail market Market accepts demand and supply bids Clears on five minute intervals Designed to also manage capacity constraints at substation More Comfort More Savings Residential energy management system Acts as a distributed agent to offer bids & respond to clearing prices Consumer sets a preference for savings versus comfort Currently being tested as part of the AEP gridsmart ARRA Demonstration in Columbus, OH

Real-Time Price / Double Auction Market Typical Unconstrained Conditions P, Price ($/MWh) P clear base Unresponsive Loads Base retail price based on PJM 5- min real- 6me market Responsive Loads Demand Curve: sorted (P, Q) bids from RTP- DA customers Feeder Supply Curve Market clears every 5-mins to ~match AC load cycle Cleared load varies with demand curve Clearing price is constant at base retail price Varies every 5- min Q, Load (MW) Q clear Feeder Capacity

Real-Time Price / Double Auction Market Typical Constrained Conditions P, Price ($/MWh) P clear Unresponsive Loads Base retail price based on PJM 5- min real- 6me market Feeder Supply Curve Responsive Loads Demand Curve: sorted (P, Q) bids from RTP- DA customers Market clears every 5-mins Cleared load is constant at feeder capacity Clearing price varies to keep load at capacity P base Feeder Capacity Q clear Q, Load (MW)

Ideal result is Decreased wholesale energy costs Peak demand limited to feeder capacity www.gridlabd.org IEEE-13 node system with 900 residential loads simulated in GridLAB-D

But what happens when including communication latency? IEEE-13 node model with 900 residential loads and controllers modeled in GridLAB-D Model was modified to work within FNCS framework An ns-3 communication network model was created (radial WIFI) EXTREME communication delays were considered

But what happens when including communication latency? Excessive communication delays during critical period caused an accounting error in auction (this was considered in Demo deployment) www.nsnam.org www.gridlabd.org As simulated in GridLAB-D and ns-3

Now let s add a transmission element Want to be able to integrate >2 simulators ns-3 GridLAB-D transmission solver P line Example: Wide Area Monitoring, Protection, and Control (WAMPAC) Controller Want to limit the power flowing through branch 34 GridLAB-D Use a price signal broadcasted to a distribution circuit to limit demand

What data is being exchanged? GridLAB-D is posting current load to a transmission substation The transmission solver is performing power flow calculations with updated load information P line Controller The control object is calculating the change in price needed GridLAB-D A new price is being broadcasted to distributed devices in GridLAB-D via ns-3

Some results Reduced demand in distribution simulator, but reflected in transmission simulator PI controller takes some time to learn the necessary price adjustment (not well tuned) Power (MW) Time (hours) But, we can see the response within GridLAB-D Reduces the demand from hour 40 through 46 Price signal is being produced in the transmission solver (this could be replaced with MATPOWER & LMPs) Price is broadcasted via ns-3 (we could look at affects of communication delays)

A few comments Simple example to demonstrate the simulation environment Any tool could be replaced with another of better value Complexity of design is up to the user We are testing multiple distribution and communication networks on a single small network A co-simulation environment can help determine the most economic means of deploying smart grid technologies, specifically in terms of communication requirements for successful system operations How much communication infrastructure do I need? What affect will latency have on my monitoring / control scheme? This will become more important as multiple applications are layered over the same communication systems

Questions?

Why a federated approach? Does not require users to provide additional models (e.g., expectation models) for co-simulation. Support both transmission and distribution level power simulators. Message delivery without incurring delays. Re-use of existing power grid and network models. Efficient synchronization algorithms to handle performance issues.

FNCS Framework Architecture " Fenix is a federated co-simulation framework: " Each simulator runs in its own process, Fenix handles the time synchronization and inter-simulator message delivery. " Federation allows all models provided by the simulators to be used in integrated simulations. Fenix Simulator API Fenix Simulator API Fenix Broker Fenix Simulator API

Relatively simple control design Simple PI control design Only used to show how the software works (does not deal with revenue, price as a signal, regulatory issues, etc.) Demand response consumers are using the same mechanism as previous use case Price is now derived as a function of the system constraints